Single-molecule analysis reveals widespread structural variation in multiple myeloma

Significance

In the last several years, we have seen significant progress toward personalized cancer genomics and therapy. Although we routinely discern and understand genomic variation at single base pair and chromosomal levels, comprehensive analysis of genome variation, particularly structural variation, remains a challenge. We present an integrated approach using optical mapping—a single-molecule, whole-genome analysis system—and DNA sequencing to comprehensively identify genomic structural variation in sequential samples from a multiple myeloma patient. Through our analysis, we have identified widespread structural variation and an increase in mutational burden with tumor progression. Our findings highlight the need to routinely incorporate structural variation analysis at many length scales to understand cancer genomes more comprehensively.

Abstract

Multiple myeloma (MM), a malignancy of plasma cells, is characterized by widespread genomic heterogeneity and, consequently, differences in disease progression and drug response. Although recent large-scale sequencing studies have greatly improved our understanding of MM genomes, our knowledge about genomic structural variation in MM is attenuated due to the limitations of commonly used sequencing approaches. In this study, we present the application of optical mapping, a single-molecule, whole-genome analysis system, to discover new structural variants in a primary MM genome. Through our analysis, we have identified and characterized widespread structural variation in this tumor genome. Additionally, we describe our efforts toward comprehensive characterization of genome structure and variation by integrating our findings from optical mapping with those from DNA sequencing-based genomic analysis. Finally, by studying this MM genome at two time points during tumor progression, we have demonstrated an increase in mutational burden with tumor progression at all length scales of variation.

Multiple myeloma (MM) is the malignancy of B lymphocytes that terminally differentiate into long-lived, antibody-producing plasma cells. Like other cancers, it is characterized by many genomic aberrations, including single nucleotide variants (SNVs) (1, 2), translocations (most notably involving the Ig heavy chain locus on chr14), and copy number changes, including aneuploidy (3). Recent large-scale sequencing studies have described widespread inter- and intra-tumor genomic heterogeneity (1, 2), clonal evolution (4, 5) and clonal tides (4) in MM. However, most of this work focuses on point mutations and large-scale copy number changes. Although the role of structural variation in normal human genome polymorphism (6, 7) and diseases (8) is widely appreciated, a comprehensive analysis of structural variation in MM is yet to be reported.

The therapeutic landscape for MM over the past decade has been transformed with the introduction of proteasome inhibitors (bortezomib, carfilzomib) and thalidomide analogs (9, 10). Consequently, patient survival rates have vastly improved (11). However, MM remains an incurable cancer, and almost all patients with symptomatic MM die of their disease because acquired drug resistance limits the efficacy of current therapies and shortens overall survival (12). Therefore, understanding the impact of contemporary treatments on MM genomic selection may provide fundamental insights for preventing and/or circumventing drug resistance through judicious use of existing therapies and/or rational design of novel agents.

To address these issues, we have used optical mapping (7, 13–19) and DNA sequencing to comprehensively characterize structural variation in a primary MM genome at two stages of tumor progression and drug response. The two stages represent a sensitive relapse (MM-S; patient responded to subsequent treatments) and a subsequent refractory relapse (MM-R; no response to any treatments) (SI Materials and Methods and Fig. 1). Optical mapping is a single-molecule system that constructs large datasets comprising ordered restriction maps (Rmaps; 1 Rmap is a restriction map of a single DNA molecule) from individual genomic DNA molecules (Fig. S1). These datasets are submitted to a computational pipeline powered by cluster computing for genome assembly (15) and discovery of structural variants (7, 14, 16, 19). The final assembly presents a relatively unbiased, long-range view of the genome, free of amplification and cloning artifacts, which supports the identification of structural variants and large-scale copy number changes. Previously, optical mapping has been used to uncover structural variation in normal (7), disease risk (17), and cancerous (18) human genomes. Here, we connect long-range structural variation findings from optical mapping with results from whole genome DNA sequencing data analysis (Fig. 1). Such analysis has enabled us to comprehensively identify somatic variation in these tumor samples across all length scales, including structural, copy number, and single nucleotide variation. Additionally, by analyzing these tumor samples at two time points during tumor progression, we have highlighted an increase in mutational burden with tumor progression.

Fig. 1.

Overview of cancer genome analysis pipeline comprising optical mapping and DNA sequencing data. Red text indicates that the method identifies somatic variation directly by comparing the tumor to the normal sample. Colored outlines highlight different variation types analyzed by integrating data from both approaches; for example, deletions from optical mapping (blue outline) were analyzed along with deletions from BreakDancer, Pindel, and CNVnator.

Results

Rmap Alignments Reveal Widespread Copy Number Changes in the MM Genome.

In any region of the genome, the total number of aligned Rmaps (depth of coverage) serves as an indicator of copy number. For somatic copy number analysis using optical mapping data, we compared the depth of coverage of both tumor samples (MM-S and MM-R) to a reference dataset (normal) using a hidden Markov model-based coverage analysis algorithm (18, 19). As a result, the tumor genomes were partitioned into low, normal, and high copy number states. This analysis is analogous to traditional hybridization or sequencing-based copy number analysis (20) because alignment of Rmaps (300–2,500 kb in length) is scored in place of probes or short sequence reads. On comparing MM-R with paired normal sample, we found widespread genomic gains and losses that spanned close to one-third of the reference genome and were generally associated with chromosomal ends (Fig. 2, ring D). A comparison of optical mapping-based copy number analysis with DNA sequencing-based copy number analysis (21) revealed that, for events greater than 500 kb, 97% of the genome was assigned concordant copy number state by both methods (Fig. 2, rings D and E). Finally, copy number changes spanning ∼172 Mb, relative to the reference genome, were observed only in the MM-R sample and not in the MM-S sample, indicating an increase in copy number changes with tumor progression (Fig. 2, sectors highlighted in yellow background color).

Fig. 2.

Circos plot of genomic variation in MM. Tracks are as follows: The outer ring represents reference chromosomes 1 through 22 in clockwise orientation (chr8 reversed; chrs X and Y excluded for clarity; numbers on the ring represent chromosomal position in Mb). Ring A: Green bars show somatic deletions shared between MM-S and MM-R samples; brown bars show somatic deletions additionally acquired in MM-R sample. Ring B: Green dots show nonsynonymous SNVs shared between MM-S and MM-R samples; brown dots show nonsynonymous SNVs additionally acquired in MM-R sample. Ring C: Density plot of loci where normal sample is heterozygous, whereas the MM-R tumor sample shows loss of heterozygosity; light green bars highlight genomic regions with copy number neutral loss of heterozygosity. Ring D: Somatic copy number (CN) changes identified from Rmap coverage analysis of MM-R vs. normal sample; low CN (blue), normal CN (gray), and high CN (red) states are shown. Ring E: Somatic CN changes from DNA sequencing data analysis of MM-R vs. normal sample; CN loss (blue), normal CN (gray), and CN gain (red) regions are shown. Ring F: Links represent structural rearrangements that coincide with and lead to CN changes identified from Rmap assembly and/or DNA sequencing data. Gray links connect translocation breakpoints; red link represents the t(11,14) translocation; blue links represent deletions; green links represent inversions. Yellow highlights (background; center to outside the circle) highlight CN changes that are observed only in the MM-R, but not the MM-S sample.

Rmap Assemblies Reconstruct the MM Genome and Characterize Structural Rearrangements.

Consensus optical maps, constructed from iterative assembly of Rmaps, generate a genome-wide scaffold providing nearly telomere-to-telomere information about the genome under study. We analyzed chimeric consensus maps, which are formed as a result of interchromosomal rearrangements or intrachromosomal rearrangements separated by at least 300 kb, and found that the location of many chimeric consensus maps coincided with copy number breakpoints (seen in Fig. 2, rings D and E). Using this analysis and integrating it with DNA sequencing-based structural variation analysis, we characterized genomic rearrangements at 31 out of 37 copy number breakpoints observed in the MM-R genome to base pair resolution. These rearrangements have been summarized in Table 1 and include unbalanced translocations, interstitial deletions, chromosomal truncations, tandem duplications, and more complex rearrangements (Fig. 2, ring F). By combining individual events, we pieced together the structure of many chromosomes, effectively generating a karyotypic representation of these chromosomes. Here, we describe the structure of two chromosomes, chr2 and chr5, in more detail.

Table 1.

Structural variants that underlie large-scale copy number changes in MM-R sample

Chr1	Loc1	Gene_Loc1	Chr2	Loc2	Gene_Loc2	Event
Chr1	46,360,804	MAST2	Chr1	117,273,548		Deletion
Chr1	160,523,168	CD84	Chr6	51,118,469		Translocation
Chr1	180,927,627		Chr9	131,339,039	SPTAN1	Translocation
Chr2	138,855,727		Chr2	162,035,600	TANK	Deletion
Chr2	208,070,175		Chr10	23,276,740	ARMC3	Translocation
Chr4	170,232,951		Chr4	182,336,593		Deletion
Chr5	70,740,000					Truncation
Chr5	78,413,500	BHMT	Chr5	96,631,200		Tandem Duplication
Chr5	134,007,729	SEC24A	Chr7	66,958,972		Translocation
Chr5	180,039,850	FLT4	Chr11	119,453,437		Translocation
Chr7	20,878,695		Chr16	78,290,600	WWOX	Translocation
Chr7	46,962,661		Chr16	5,998,555		Translocation
Chr7	67,099,767		Chr7	67,262,743		Inverse Duplication
Chr7	120,756,417	CPED1	Chr12	20,883,779	SLCO1C1	Translocation
Chr8	36,284,561		Chr9	32,013,790		Translocation
Chr8	39,393,625		Chr8	41,441,031	AGPAT6	Deletion
Chr11	69,453,552		Chr14	106,107,045		Translocation
Chr14	39,296,961	LINC00639	Chr21	29,006,964	MIR5009	Translocation
Chr14	48,211,973		Chr14	52,920,149	TXNDC16	Inversion
Chr15	30,873,871	ULK4P1, ULK4P2	Chr15	31,189,559		Deletion
Chr17	2,023,044	SMG6	Chr17	13,892,185		Deletion
Chr17	13,892,185		Chr17	15,895,185	ZSWIM7	Inversion
Chr17	15,895,213	ZSWIM7	Chr17	16,236,159		Deletion
Chr17	44,041,355	MAPT	Chr19	58,412,876		Translocation
Chr18	52,209,201		Chr9	Past End		Translocation

Detailed structure of chr2.

Chr2 presents with two regions of copy number loss, a 23 Mb region spanning from 138.85 Mb to 162.03 Mb and another 35 Mb region spanning from 208.07 to 243.19 Mb (Fig. 2). Consensus maps revealed an interstitial deletion that explains the loss of 138.85–162.03 Mb (Fig. 3A). Additionally, we identified an unbalanced translocation between chr2 and chr10, t(2;10)(q33.3;p12.2), which explains loss of the 35 Mb region on chr 2 (208.07–243.19 Mb) and amplification of the 23 Mb region on chr10 (23.2 Mb to start) (Fig. 3B).

Fig. 3.

Genomic reconstruction of chr2 in MM-R sample. Chr2 presents with an interstitial deletion [138.85–162.03 Mb; del(2)(q22.1;q24.2)] and an unbalanced translocation with chr10 [208.07 Mb; t(2,10)(q33.3;p12.2)] in MM-R sample. Chimeric consensus maps, constructed from Rmap assembly, reveal breakpoints (red lines) and span ∼5.75 Mb at each breakpoint to elucidate long-range genomic structure at the deletion (A) and the translocation (B) events. The length of each restriction fragment is proportional to its size (in kb). Spanning each breakpoint, fluorescence micrographs of six BamHI digested DNA molecules are shown; each DNA molecule comprises a series of daughter restriction fragments, which appear as dashed white lines. These single-molecule Rmaps, along with others, were assembled to generate the consensus maps.

Detailed structure of chr5.

Chr5 presents with two regions of copy number loss, 70.74–78.41 Mb and 96.63–134.01 Mb (Fig. 2). From Rmap assembly, we observed a chromosomal end at 70.74 Mb, indicating that the region after this breakpoint was lost via truncation in one copy (Fig. 4A). Furthermore, we identified tandem duplication of an 18 Mb long region spanning 78.41 Mb to 96.63 Mb, which explains copy number neutral loss of heterozygosity in this region (Fig. 4B). We were not able to explain normal copy number for the region spanning 134.01 Mb to the end of chr5. Finally, we observed an unbalanced translocation between chr5 and chr11, t(5;11)(q35.3;q23.3), which is associated with loss of ∼900 kb at q-ter of chr5 (180.03–180.91 Mb) and amplification of ∼15.5 Mb from q-ter of chr11 (119.45–135.00 Mb) (Fig. 4C). It is interesting to note that the 70.74 Mb truncation breakpoint overlaps a segmental duplication and that the 96.63 Mb tandem duplication breakpoint overlaps a LINE repeat, thereby obscuring these regions to full analysis using DNA sequencing data alone. We were able to resolve and understand the highly rearranged structure of chr5 by integrating information from Rmap assembly, copy number analysis, and DNA sequencing data.

Fig. 4.

Genomic reconstruction of chr5 in MM-R sample. Chr5 presents with a truncation (A), a tandem duplication (B), and an unbalanced translocation with chr11 (C) in the MM-R sample. (A) Alignment of consensus map (OM) to in silico reference map (chr5, 70.5–71 Mb) is shown, and indicates truncation. The reference map is annotated with chromosomal band, RefSeq genes, and segmental duplications (top to bottom) from the University of California, Santa Cruz (UCSC) genome browser. (B) Alignment of consensus map (OM) to in silico reference maps (chr5, 78.15–79 Mb, red outline; and chr5, 96.0–96.8 Mb, dark gray outline) is shown, highlighting tandem duplication of 78.4 to 96.6 Mb region. (C) Alignment of consensus map (OM) to in silico reference maps (chr5, 179.47–180.48 Mb; and chr11, 118.83–119.84 Mb) is presented. The topmost and bottommost tracks annotate reference maps from chr5 and chr11 with copy number profiles obtained from DNA sequencing data analysis.

Structural Rearrangements Explain Many Canonical and Noncanonical MM Copy Number Changes.

The copy number changes observed in these tumor samples represent canonical and noncanonical MM genomic losses/gains and examples of some of these changes, drawn from Fig. 2, are described here along with underlying structural rearrangements.

Concerning 1p loss, we identified a 72 Mb long deletion, del(1)(p34.1;p13.1), on the short arm of chr1. The other chr1 allele presented a 112 kb deletion at 1p32.3, encompassing the genes FAF1 (FAS [TNFRSF6]-associated factor 1) and CDKN2C (Fig. S2), leading to homozygous loss of FAF1 and CDKN2C. The deletion of negative cell cycle regulator CDKN2C is known to be a key early aberration in MM (22). Concerning chr11 and chr14, the t(11;14)(q13;q32) translocation involves the Ig heavy chain locus at 14q32 and the cyclin CCND1 at 11q13, and leads to overexpression of CCND1. The chr11 breakpoint is known to vary from 1 kb to 1 Mb upstream of CCND1 (23). We identified t(11;14) in both tumor samples (Fig. S3). Consistent with previous reports, the breakpoint on chr11 is ∼2 kb upstream of CCND1. Concerning chr13, we identified loss of one copy of chr13 from Rmap alignment based coverage analysis. Concerning chr17, deletions involving TP53 locus occur in ∼10% of untreated MM patients (24). We observed an ∼12 Mb deletion in the 17p region, which includes the gene TP53. This deletion forms part of a complex genomic rearrangement, where two genomic segments (2 Mb to 13.89 Mb and 15.89 Mb to 16.23 Mb) are deleted and the intervening region (13.89 Mb to 15.89 Mb) is inserted in an inverted orientation (Fig. S4). Concerning chr14, we identified a 4.71 Mb long inversion at the 14q21.3-q22.1 locus (Fig. S5). Other rearrangements, for example, a translocation between chr17 and chr19, t(17;19)(q21.31;19q13.43), led to amplification of ∼37 Mb at q-ter of chr17 and loss of ∼1Mb at q-ter of chr19 (Fig. S6). In another event, an ∼26 Mb long region from q-ter of chr18 was found to be amplified in a fusion at the end of chr9 (Fig. S7).

Analysis of Deletions from Optical Mapping and DNA Sequencing.

Using an automated pipeline that identifies structural variants, the consensus maps from Rmap assemblies were compared via alignment to in silico restriction maps generated from the human reference sequence. Such a comparison identifies intrachromosomal structural variants like deletions and insertions, ranging in size from 3 kb to ∼300 kb.

Using this pipeline, we identified 139, 149, and 176 deletions in normal, MM-S, and MM-R samples, respectively, ranging from ∼3 kb to 180 kb in size, with mean size ∼10 kb and median size ∼6 kb (Fig. S8A and Dataset S1, Table S1). We compared the deletions from optical mapping to those from DNA sequencing data analyzed using read-pair–based (25), split-read–based (26), and read-depth–based (27) approaches. Close to 80% of the optical mapping deletions were validated by one or more sequencing-based methods (Fig. S8C). We further analyzed the deletions identified only by optical mapping in MM-R sample and found that a majority of them overlapped with segmental duplications (13/38) or repeat DNA sequences (11/38). Overall, this analysis points to high accuracy of deletion calling and the ability to identify variation in duplication/repeat rich regions using optical mapping.

Previous analysis of normal human samples using optical mapping has revealed a large number of structural polymorphisms in normal genomes (7). Accordingly, most of these deletions, as evidenced by 139 deletions observed in the normal sample, are germline and need to be filtered out to identify somatic deletions. Using a combination of optical mapping and DNA sequencing analysis, we identified many somatic deletions that we divided into two categories: shared between MM-S and MM-R samples (Fig. 2, ring A, green bars) and acquired additionally in the MM-R sample (Fig. 2, ring A, brown bars). Based on the approaches used, we partitioned these deletions into two size ranges: 10–400 bp and >400 bp (Dataset S1, Table S2).

Somatic deletions shared between MM-S and MM-R samples.

We identified 38 somatic deletions larger than 400 bp, which range in size from 497 bp to 192 kb (Dataset S1, Table S3A). Of these findings, 10 deletions overlap exons. Furthermore, we identified 6 small deletions (10–400 bp) that overlap exons (Dataset S1, Table S3B). Among these deletions, we noticed a 73 bp deletion associated with a TP53 exon. As discussed before, MM-S and MM-R samples also show del(17p), which indicates that both copies of TP53 are inactivated.

Somatic deletions acquired additionally in MM-R sample.

We identified another 27 somatic deletions larger than 400 bp, which range in size from 445 bp to 290 kb and were additionally acquired in the MM-R sample (Dataset S1, Table S4A). Of these 27, 7 deletions overlap exons. Furthermore, we identified 2 small deletions (10–400 bp) that overlap exons (Dataset S1, Table S4B). Similarly to large-scale copy number changes, an increasing number of somatic deletions highlights increasing mutational burden with tumor progression.

Analysis of Insertions from Optical Mapping and DNA Sequencing.

Using the automated variation calling pipeline, as described above for deletions, we identified 450, 428, and 384 insertions in normal, MM-S, and MM-R samples, respectively, ranging from ∼3 kb to 367.5 kb in size, with mean size ∼10 kb and median size ∼5 kb (Fig. S8B and Dataset S1, Table S1). Although deletions are relatively straightforward to identify using sequencing-based approaches, identifying insertions can be problematic. We compared the insertions from optical mapping to insertions identified using DNA sequencing data analysis approaches Pindel (26) and CNVnator (27). Pindel generates insertion breakpoints with no information about the length or structure of the insertions whereas CNVnator generates a list of duplications with no structural or contextual information. As expected, a much smaller percentage (24–40%) of optical mapping insertions had an overlapping Pindel, or CNVnator call (Fig. S8D). Finally, we did not discover any somatic insertions from our analysis.

SNP/SNV Analysis from DNA Sequencing Data.

Commonly associated SNPs.

Of the eight inherited susceptibility markers that have been identified by genome-wide association studies (GWAS) in MM (28–30), we found seven in our tumor samples (Dataset S1, Table S5). Among these markers is an SNP in gene CCND1 (c.870G > A, rs9344) at 11q13.3 that has been implicated as a risk factor for t(11;14)(q13;q32) MM (29).

SNV analysis reveals increasing mutational burden with tumor progression.

We identified 10,224 and 13,511 SNVs in MM-S and MM-R tumor samples, respectively, compared to the normal sample. This analysis yielded tumor-specific point mutation rates of 3.53 and 4.52 per million bases for MM-S and MM-R samples, respectively, indicating an increase in mutational burden with tumor progression. Although a bit higher than the average tumor-specific point mutation rate from the Multiple Myeloma Research Consortium (MMRC) cohort (2.9 per million bases) (1), it can be explained by higher sequencing depth or increased incidence of somatic mutations in our samples. No SNVs were shared between our work, the MMRC cohort (1), and a previous study by Egan et al. (4).

Functional annotation of SNVs and further analysis of nonsynonymous SNVs revealed that 60 such SNVs were shared between both tumor samples (Fig. 2, ring B, green dots). The MM-R sample acquired an additional 41 SNVs (Fig. 2, ring B, brown dots), corroborating the increase in mutational burden with tumor progression (Dataset S1, Tables S6 and S7). Consistent with previous findings from primary MM samples (31), we did not find any PSMB5 proteasomal mutations.

SNP analysis identifies regions with copy number neutral loss of heterozygosity.

From SNP/SNV analysis, we identified genomic loci that were characterized as heterozygous in the normal sample but homozygous in the tumor samples, thereby reflecting loss of heterozygosity. Upon plotting the density of such loci across the genome (Fig. 2, ring C), we observed, expectedly, that many regions with increased density overlapped with copy number losses. However, four regions spanning a total of 215 Mb on q arms of chr1, chr5, and chr14 presented with copy number neutral loss of heterozygosity in the MM-R sample (Fig. 2, ring C, light green bars).

Discussion

Comparison with Existing Genomic Methods.

Traditionally, chromosomal karyotyping, fluorescence in situ hybridization (FISH), and, more recently, hybridization and sequencing based technologies have been used to study clinical MM samples. However, they have certain limitations. FISH can address only known locations of structural variation and thus precludes comprehensive discovery. Karyotyping, on the other hand, has limited resolution and requires actively dividing cells, which is generally an issue for MM tumor cells because they are known to have a low proliferative index (32). Genome-wide hybridization technologies do not directly provide structural information about genome structure whereas sequencing-based technologies are either limited by cost to exome analysis or have generally been used to study the patterns of single nucleotide variation (1, 2). With optical mapping, we use primary tumor samples and therefore are able to identify structural variation representative of underlying cell population free of hybridization, library creation, or amplification artifacts. Also, the resolution of structural variants (∼3 kb) effectively generates a high resolution genomic karyotype.

Single Nucleotide vs. Structural Variation.

Our work did not find any common SNVs shared between this study, a previous study (4), and the MMRC cohort (1). The lack of common SNVs and widespread distribution of structural variation indicate that structural variation might play a larger than appreciated role in explaining MM pathogenesis/drug resistance and warrants the need for a population study of structural variation in MM and other cancers.

Implications for MM Biology.

The analysis presented here, although limited to a single individual, offers unique biological insights that may translate into novel therapeutic strategies. Individual genes mutated in both MM-S and MM-R samples may represent clonal drivers associated with core mechanisms of MM oncogenesis and/or acquisition of drug resistance. These events constitute legitimate therapeutic targets, particularly in light of recent data demonstrating the limits, and potential dangers, associated with targeting subclonal mutations (e.g., mutant BRAF) (2). Several events have been established previously as important progression factors in myeloma pathogenesis (e.g., TP53 and CDKN2C loss) (22, 24). PIK3R1 aberrations underscore the importance of the PI3K pathway activation in MM (33). IGF2BP2 mutations affect insulin growth factor-2 (IGF-2) translation and thus growth control through an IGFR-controlled pathway that is currently the focus of intense preclinical development in myeloma (34). Other genes are less well studied in MM and may provide novel potential targets or regulatory pathways. ELL is an essential cofactor of the superelongation complex (SEC), a central node of transcriptional elongation checkpoint control and a key mutational target in myeloid and mixed-lineage leukemias (35). TCL1 is a regulator of apoptosis implicated in B-chronic lymphocytic leukemia as well as T-cell lymphomas (36). ASXL3 belongs to a family of epigenetic regulators whose genetic loss has been implicated in the myelodysplastic syndromes and myeloid leukemias (37). CAMK2D is a subunit of calmodulin-dependent kinase II, an essential regulator of Ca²⁺-dependent signal transduction. Interestingly, prior studies have implicated calmodulin-dependent pathways in proteasome inhibitor-resistant, constitutive NFκB activity in MM (38, 39). Our analysis has therefore pinpointed potential targets that merit validation in larger cohorts of MM patients before functional experimentation using appropriate model systems.

Intriguingly, genes uniquely affected in the MM-R sample encode cell cycle regulators (CCNG2) and mitotic checkpoint genes (ZWILCH), as well as a transcription factor (MYBL1) expressed specifically in centroblasts (40), a putative preplasmablastic cell-of-origin for MM within the germinal center reaction (24). These findings raise the possibility that disease progression and/or acquisition of drug resistance in MM may be associated with plasmacytic maturation arrest or dedifferentiation to earlier stages of B-cell ontogenesis. Thus, effective management of end-stage MM may necessitate approaches that promote mitotic quiescence or cell cycle exit and terminal plasmacytic differentiation.

Conclusions

Using optical mapping and DNA sequencing, we have characterized genomic variation in sequential samples obtained from an MM patient with progressive disease. Although these platforms are revealing genome analysis systems on their own, their discernment of genomic variation is complementary. Combining the unique advantages of these systems has potentiated the comprehensive understanding of genomic structure in this tumor genome and has revealed widespread variation—across the entire length spectrum of variation.

Materials and Methods

Case History, Study Design, Data Acquisition, and Rmap Construction and Assembly.

This study was approved by the Institutional Review Board at the University of Wisconsin–Madison in accordance with the Declaration of Helsinki. DNA samples were prepared from purified CD138⁽⁺⁾ plasma cells (MM-S and MM-R sample) and paired cultured stromal cells (normal) from a 58-y-old male MM patient with International Staging System (ISS) Stage IIIb disease, who had been treated with combinations of bortezomib, dexamethasone, lenalidomide, cyclophosphamide, and tandem autologous stem cell transplants at different stages of tumor progression (SI Materials and Methods). Large Rmap datasets, comprising ∼2 million Rmaps for each sample (normal, MM-S, MM-R), were constructed using BamHI restriction endonuclease. Rmap datasets thus obtained were submitted to our iterative assembly pipeline (7, 14–16), which constructs contigs and associated consensus maps from alignment and assembly of individual Rmaps and is used to identify structural and copy number variation. An overview of Rmap collection and assembly output is provided in Dataset S1, Table S8. The contig assemblies indicate almost complete (>99.5%; merged contig average size is >31 Mb) coverage of sequence scaffolds from human reference sequence [National Center for Biotechnology Information (NCBI) Build 37].

DNA Sequencing.

We obtained 100×2 bp Illumina paired end sequencing data with properly paired mean coverage depth of 59×, 68×, and 92× for normal, MM-S, and MM-R samples, respectively, with an insert size around 250 bp (Dataset S1, Table S8).

Please refer to SI Materials and Methods for more information on materials and methods.